Nor for perssnlichen Gebrouch Intemational Journal of Food Microbiclogy International Journal of Food Microbiology $3 (1999) 33-31 www.elsevier.nl/locate/ijfoodmicro Improved screening procedure for biogenic amine production by lactic acid bacteria ae Sara Bover-Cid*™, Wilhelm Heinrich Holzapfel” “Department of Nutrition and Food Scier e-CERTA. Faculty of Pharmacy, Uni E-OSO28 Barcelona, Spain sity of Barcelona, Av, Joan XXIL sin" Crseereye “Institute of Hygiene and Tosicalogy. BEE, Haid-und-New-Sie, 2, D-TO131 Kurlsrule, Germany ya panna Received 20 May 1999: accepted 7 September 19) one Abstract An improved sereening pkite method for the detection of amino acid decarboxylase-positive microorganisms (especially lactic acid bacteria) was developed. The suitability and detection level of the designed medium were quantitatively evaluated by confirmation of ami pacity using an HPLC procedure, The potential to produce the biogenic amines (BA) rumine, histamine, putrescine, and cadaverine, was investigated in a wide number of lactic acid bacteria (LAB) of different n. including starter cultures, protective cultures, type strains and strains isolated from different food products. Also, Several strains of Enterobacteriaceae were examined. Modilications to previously described methods included lowering glucose and sodium chloride concentrations, and increasing the buffer effect with calcium carbonate and potassium Phosphate. In addition, pyridoxal-S-phosphate was included as a codeearboxylase factor for its enhancing effect on the amino acid decarboxylase activity. The screening plate method showed a good correlation with the chemical analysis and due to its simplicity it is presented as a suitable and sensitive method to investigate the capacity of biogenic amine production by LAB. Tyramine was the main amine formed by the LAB strains investigated. Enterococci, camobacteria and some strains of lactobacilli, particularly of Lb. curvatus, Lb. brevis and Lb. buchneri, were the most intensive tyramine formers. Several Strains of lactobacilli, Leuconostoc spp.. Weissella spp. and pediococci did not show any, potential to produce amines. Enterobacteriaceae were associated with cadaverine and putrescine formation. No significant histamine production could be detected for any of the strains tested. © 1999 Elsevier Science BY. All rights reserved, Keswords: Decarboxylase activity: Biogenic amines (BAY: Lactic acid bacteria (LAB); Enterobacteria 1. Introduction pounds which occur in different kinds of food. such : as fishery products, cheese, wine. beer. dry sausages Biogenic amines (BA) are organic basic com- and other fermented foods (ten Brink et al., 1990; Halasz et al., 1994). ~ hi faxes Several toxicological problems resulting from the Longo ne NOE Tels + SESIODSI: fa: +95 Food containing relatively high levels of E-mail address: bover@farmacia.far.ub.es (S. Bover-Cid) BA have been reviewed (i.e., ten Brink et al., 1990; 168-1605/99/S - see front matter © 1999 Elsevier Science BY. All rights reserved. PUL: $0168-1605(99)00152-Xa S. Bover-Cid, WH. Holzapfel | International Journal of Food Microbiology 33 (1999) 33-21 ‘Mariné-Font et al., 1995). Histamine and tyramine have been the most studied BA due to the toxicologi- cal effects derived from their vasoactive and psycho- active properties. Histamine has been recognised as the causative agent of scombroid poisoning (his- taminic intoxication), whereas tyramine has been related to food-induced migraines and hypertensive crisis in patients under antidepressive treatment with mono-amine oxidase inhibitor (MAOI) drugs. Al- cohol and other BA. such as the diamines. putrescine and cadaverine, may boost the toxicity of the above amines. In addition, diamines are known to be potential precursors of carcinogenic nitrosamines, especially when nitrosable agents are present in food. Apart from these toxicological aspects, BA are of concer in relation to food hygiene. The occurrence of relatively high levels of certain BA has been reported as indicators of a deterioration process and/ or defective elaboration (Karmas, 1981; Vidal-Carou et al., 1990). In foods, BA are mainly gencrated by decarboxyl- ation of the corresponding amino acids through substrate-specific enzymes of the microorganisms present in the food (ten Brink et al., 1990), The capability of BA formation has been described for several groups of microorganisms, mainly En- terobacteriaceae, Pseudomonas spp., enterococci and some other LAB (Halisz et al., 1994). The occur- rence and distribution of amino acid decarboxylase activity within different genera and species of the Enterobacteriaceae, mainly isolated from fish, has been extensively reported. By contrast, studies on the BA formation by LAB are still relatively scarce. Several qualitative and quantitative methods to determine BA production by microorganisms have been described. Most of the screening procedures generally involve the use of a differential medium containing a pH indicator, such as bromocresol purple (Moeller, 1954; Niven et al, 1981; Choudhury et al.. 1990). Usually, the media consist of a basal composition (peptone, yeast or meat extract, salt and/or glucose) to which the precursor amino acids are added (Table 1). A positive result is indicated by a change of the medium colour to Table 1 Composition (4) of decarboxylase media according to the literature snd! of the improved medium Component Moeller Niven et al Choudhury etal. Joosten and Northold Maja Improved (19s) 9gt) 1990) (1989) cs) medium Tryptone 0s os 0s Os 05 Os Yeast extract 0s 4 os 0s o4 05 Meat extract - - - 2 08 Os Sodium chloride 0.8 os os os - 025 Glucose - - 0.05 on - 04s Tween KO" - - - 0.05 0.05 on - - - 0.02 oor 002 - - a 0.005 0.005 0.005 - - a 0.004 0.008 0.004 - 5 2 5 - 02 Thiamine - - - . - 0.001 K,PO, - 7 _ a - 02 Caco, on ou oor oor oon Pyridoxal-s- 00s . . S 5 0.005 Phosphate os Amino acid 05-10 27 (His, (0.008 for Tyr) 2.0 20 10 Bromeresol 6.001 0.006, os 0,006 0.006 0.006 purple Cresol red 0.0008 6 5 7 - Agar - 2 2 7 2 pH 60 38 50 53 53 Application Emerobacteria Enterobacteria Lb. buchneri Lactobacilli LAB from LAB and from fish 0. meni from cheese meat products enterobacteriS. Bover-Cid, WH. Holzapfel | Imemational Journal of Food Microbiology purple in response of the indicator to a pH shift. The pH change is dependent of the production of the more alkaline BA from the amino acids initially included in the medium. Such indicator media have been used in several studies for the detection of amine-producing en- terobacteria in fish and fish products (Chen et al.. 1989). Yet. some reports (Baranowski, 1985; Roig- Sagués et al., 1997) have described the occurrence of false-positive reactions. due to the formation of other alkaline compounds. False-negative responses as a result of the fermentative activity of some bacteria which produce acid along with BA, constitute another major problem when screening LAB. In addition, some microorganisms, such as fastidious ly simple s reported by Joosten and Northold (1989) and Maijala (1993) in order to adapt the method to different purposes (Table 1). Joosten anid Northold (1989) added some metal sulphates (Mg, Mn, and Fe) and Tween 80 to the Niven’s medium (Niven et al., 1981) to enhance the growth of lactobacilli isolated from cheese, According to these authors, the contrast of the colour surrounding the colonies was improved by lowering the initial pH of the medium to 5.0 and by addin; glucose, Optimal glucose concentration was 0.1%, as higher concentrations led to excessive lactic acid production, thereby counteracting the purple colora- tion. Tyrosine has low solubility (0.45 g/I at 25°C) and at the 2% concentration used, plates were not translucent and the characteristic colour change could not be used to detect tyramine producers. However, tyrosine-decarboxylating bacteria were surrounded by a clear halo, resulting from the disappearance of tyrosine precipitate. In spite of these modifications, Maijala (1993) reported that some non-starter LAB showed only weak growth Fesponse and were unable to produce a positive reaction. In order to facilitate the growth of meat LAB, decarboxylase agar of Joosten and Northold was modified by adding Lab-Lemco powder (meat extract) and omitting NaCl. According to Maijala (1993), glucose should be omitted because it may inhibit the appearance of the positive colour by Causing a pH decrease as it is fermented to lactic acid by LAB. Although these authors reported the suitability of 53 (1999) 33-44 5 the media to determine BA production by LAB, only few positive results were detected from ™a "wide number of LAB tested, either due 'fo' insufficient growth or no change of colour. Therefore:: the objectives of this study was to develop an improved screening method for the detection ofvamino' acid decarboxylase-positive microorganisms'*(espectally LAB). In addition. the suitability of the"dés{gne medium was evaluated by confirmation of the titative amine-for assay. The potential to produce BA ((yramine, histamine, putrescine and cadaverine) was Mavest- cultures, protes clues, type straing and! bls isolated from different food products, 2, Material and methods 2.1. Bacterial cultures tobacilli, lactococci, enterococci and carn were used in this study. The strains were,from the ATCC (American Type Culture Collection), BAFF (Bundesandtalt fiir Fleischforschung, »Kulmbach, Germany), BFE (Federal Research Center for Nutri- tion, Karlsruhe, Germany), CTC (Meat Technology Center-IRTA, Girona, Spain), DNB (Department of Nutrition and Food Science, University of Barcelona, Spain), DSM (German Collection of -Microorga- nisms, Braunschweig, Germany), and“ NCFB °(Na- tional Collection of Food Bacteria, Reading, UK) collections. Also, 17 enterobacteriaceac strains iso- lated from fermented sausages were tested? Summa- rising information on the number of strains of each species studied is given in Table 2.~: 2.2, Activation of microbial cultures In order to promote the enzyme induction before the actual screening test, LAB strains were subcul- tured 5 to 10 times in MRS broth (110661 Merck, Darmstadt, Germany), while Enterobacteriaceae strains were pregrown in Nutrient Standard broth (107882 Merck, Darmstadt, Germany), both con- taining 0.1% of each precursor amino acid (all from Merck, Darmstadt, Germany), comprising tyrosine free base, histidine monohydrochloride, omithine36 S. Bover-Cid, WH. Holzapfel | Imemational Journal of Food Microbiology 53 (1999) 33241 Table 2 List of species assayed using the improved decarboxylase screening medium: the number of strains positive for at least one BA is indicated ‘with respect to the total number tested Species Strains ‘Species Strains Lb. acidophilus 25 : Lb. alimentarius ont Camobacterium divergens 10/10 Lb. amslovorus 0/2 Carnobacterium gallinarum " Lb. animals: ont Carnobacterium mobile on Lb. bavaricus 24 Carnobacterium piscicola 144 Lb. bifermenans w" Lb, brevis, 3I4 Enterococcus durans wm Lb. brevis /buchneri ont Enterococcus faecalis Isis Lb, buchneri wt Enterococcus faecium 1or10 Lb. bulgaricus ma Lb, casei mm Leuconostoc carasum 2 Lb, cellobiosus on Leuc, mesenteroides ssp. mesenteroides uw Up, curva Ras Lee, mesenteroides ssp. cremoris on Us, delbrueckii ssp. bulgaricus 22 Leuc, mesenteroides ssp. dextranicus ont Ub, delbrueckii ssp. lactis, on Leuc. mesenteroides on Lb, fermentum on Ub, passeri on enocaceus enti on Lb, helveticus on Lb. hilgardit on Weissella confusr on Ub. imestinalis: on Weissella viridescens on Lb, jolnseni Ws Lb. paracasei ssp. paracasei 16 Pedivcoccus acidilactiet on Lb, puracasei ssp. tolerans on Pediocoecns pemosacens on Lb, plamarum 116 Lb, reuteri on Citrobacter freundii uw Lb, rhamnosus os Enterobacter cloucue 22 Lb, sakei 27 Klebsiella axswca m7 Lb, salivarius ssp, salivinus on Proneus vulgaris os Lactobacillus sp. ang Serratia liquefaciens ais Serratia marcescens wn Lactococcus luetis ssp. lactis " Serratia sp. 22 Enterobacteriaceae ut Streptococcus theemaphilus 2 monohydrochloride and lysine monohydrochloride. in addition to supplementation with 0.005% of pyridoxal-S-phosphate, 23. Screen 1g medium The medium designed was based on the Joosten and Northold (1989) medium and also on the MRS agar described by De Man et al. (1960) for the cultivation of LAB. The composition of the medium is shown in Table 1. in comparison with other decarboxylation media suggested earlier. Modifica- tions to the Joosten and Northold medium consisted in reducing glucose concentration to 0.05% in order to avoid excessive acid production that might coun- teract the pH increase by the BA formation, Also, salt concentration was slightly lower, Metal sul- phates, such as Mg (0.02%), Mn (0.005% ), and (0.004%), Tween 80 (0.1%), and ammonium citrate (0.2%) were included to enhance the growth of LAB, Since some LAB strains (especially heterofer- mentative LAB) require thiamine to grow, 0.001% of this vitamin was added. In order to improve the buffer effect and to neutralise the acid produced, 0.2% of di-potassium phosphate and 0.01% of calcium carbonate were added. Also, pyridoxal phosphate was included to the decarboxylase medium (at 0.005% ) since its presence as a cofactor for the decarboxylation reaction has a strong enhanc- ing effect on the amino acid decarboxylase actiIGale. 1946; Recsei et al.. 1985), The concentration of each amino acid was 1%. Bromocresol purple was used as pH indicator at 0.006%. The pH was adjusted to 5.3 and the medium was autoclaved for 10 min at 121°C to avoid excessive hydrolysis of the agar at the low pH (Niven et al. 1981). All strains were streaked in duplicate on the decarboxylase medium plates with and without amino acids (as control) and were incubated for 4 days at 37°C, under aerobic and anaerobic conditions (MK3 Anaerobic Work Station, dw Scientitic. Mein- trup-Labortechnic, Linden, Germany) in parallel. 24. Confirmation of biogenic amine formation Confirmation of the amine-forming capacity of each microbial culture was performed through a qualitative and quantitative chemical analysis of the BA (tyramine, histamine, putrescine and cadaverine) potentially formed in the fermenting broth. The strains, previously cultured in the MRS or Nutrient Standard broth (with the precursor amino acids and pytidoxal-S-phosphate added), were inoculated at 0.1% imo a decarboxylase broth, formulated as the screening medium but without and containing 0.5% of tyrosine, 0.25% of histidine, ornithine, and lysine. The solubility problem of tyrosine was solved using tyrosine di-sodium salt, with a solubility of 30 g/l, thus achieving a’ constant concentration of available tyrosine in the medium in all the tubes. After incubating at 37°C for 4 days under aerobic and anaerobic conditions, 2 mi of fermenting broth Were centrifuged (12 000 rpm/S min), after which 1 ml of supernatant was extracted with | ml of 0.1 N HCL. It was centrifuged again (12000 rpm/5 min), filtered through 0.45 um (Millipore Corp.. Bedford, MA, USA) and stored at — 20°C until BA analysis. * Samples were analysed for the presence of BA by high-performance liquid chromatography (HPLC System from Waters Chromatography, Milford, MA) According to the method described by Hemdndez- dover et al. (1996). The method is an ion-pair chromatography procedure based on the formation of fon-pairs between BA and octanesulfonic acid pres- €nt in the mobile phase. Their separation as neutral Compounds is carried out through a reverse phase column (Nova Pack C18, Waters). When BA are already separated. a post-column derivatization with e-phthalaldehyde (OPA) in presence of 2-mercapto- ethanol leads t0 a high recovery because’ ofta’phoet = time between the formation of the ‘unsabietdesves= uves and their detection. Detection iiwa BBy’ fluorimetry with the excitation wavelength $Y 340'hix and the emission wavelength at 445 nm:The identi cation of the BA in the samples is mide ibyithé:-- retention times according to the standaldgtethe - quantification is carried out by an extetnal ‘standard procedure through the calculation of -a7ealfbratlGa curve with standard solutions at differents trations ranging from 0.5 to 10 ppm. Chemicals All_ microbiological media and chemicaly§ a from Merck (Darmstadt, Germany) except for BA’ standards that were from Sigma (St. Louis!!MO, USA), colle ete eran ’sdt” MAL 3. Results and discussion ke roadway 3.1. Screening of decarboxylase activity ‘e.tinines larity Table 2 shows the number of positive strains from the toil number of strains tested by the sereening. procedure. Positive reactions, both on decarboxylase plates or in broth, were recorded when a purple colour occurred or tyrosine precipitate disappeared around the colonies or in the decarboxylase broth, respectively. From all the strains assayed and cone firmed by HPLC analysis, no false-positive reaction was observed. On the other hand, three tyramine. producing strains of lactobacilli (Lb. “curvanus CTC435, Lb. sakei CTC430, and Lb. curvatus CTC371) gave a negative’ response with the screen- ing procedure. However, these false negatives were shown to be weak amine formers, producing only 18, 62 and 302 mg/l of tyramine in the medium, respectively. These concentrations were 100 low to cause the pH shift and the concomitant colour change. especially when considering the acid pro- duction of these LAB strains. Neither were they able to decarboxylate tyrosine to an extent at which a clear halo in the medium could be observed. Limited information on the requirements and factors affecting amino acid decarboxylase capacity of microorganisms has been published so far. It seems important that the optimum pH to screen for38 S. Bover-Cid, WH. Holzapfel | International Journal of Food Microbiology 53 (1999) 33-41 BA production should be slightly acid (5.05.5) in order to enhance the synthesis and activity of amino acid decarboxylase enzymes. It also appears practic- able to use a pH which mimics the original food system or the product from which the organism has been isolated. The low pH may- inhibit bacterial growth to some extent, especially in the case of Enterobacteriaceae (Niven et al., 1981). However, LAB are generally acid tolerant, and should grow sufficiently at pH around 5 used in the screening procedure. Utilisation of glucose by bacteria could play a role in lowering the medium pH to a value more favour- able level for BA production. Furthermore, Gale (1946) pointed out that BA are generally produced only when fermentable carbohydrate is present in the growth medium. The presence of glucose as energy source allows the medium to support the growth and the metabolic activity of fastidious bacteria such as LAB. As is shown in Table 1, the earliest known decarboxylase differential medium of Moeller (1954) contained pyridoxal as cofactor of amino acid de- carboxylation, None of the later modifications was formulated to include this component. Although some authors supplemented the medium with pyridoxine (Frank et al. 1985), it seems that the only active form of this B vitamin as codecarboxylase factor is the o-phosphorylated pyridoxal (Gale, 1946). The presence of pyridoxal-S-phosphate has a strong enhancing effect on the decarboxylase activity expression (Gule, 1946; Reesei et al., 1985) and it, therefore, appears to be an important ingredient of the screening medium, In addi thors (Gale, 1946; ski, 1985) have reported the inducible characteristic of amino acid decarboxylase enzymes. It seems that there is an adaptive formation of these enzymes in response to the presence of the specific substrate as well as to specific conditions during growth (i.e, acid environment). Therefore, their induction may be promoted by precultivation of a strain in a medium containing the amino acids and the cofactor. According 10 Moeller (1954), strains of En- terobacteriaceae, must be protected from air during incubation in order to avoid false alkalisation at the surface of the medium. In the present work, no appreciable differences in the reaction under aerobic or anaerobic conditions were found among the LAB on, Som assayed. Yet, for Enterobacteriaceae strains, aerobic conditions always seemed to yield false-positive reactions, even on plates without any amino acid. Baranowski (1985) pointed out the importance of streaking microbial cultures on plates formulated without amino acids parallel to the screening plates in order to detect potential false-positive responses. Other pH indicators, such as bromocresol green, chlorophenol red and combinations of both, were also tested, These indicators show the colour change at lower pH than bromocresol purple, and thus might be useful to detect BA formation by the acid producing LAB, However, a high number of false positives was found (data not shown) when using the above indicators. Chen et al. (1989) also observed higher false positive results using chlorophenol red than bromocresol purple. ‘The improved decarboxylase medium sufficiently supported the growth of the strains investigated. The simplicity of the method, the easy recognition of positive reactions, and the good correlation with the chemical analysis, suggest the suitability of the present medium as a superior screening method to investigate the capacity of LAB to produce BA, 3.2. Biogenic amine production by bacteria Biogenic amine production has been most exten- sively studied with respect to histamine and tyramine, probably the two most important BA of bacterial origin in food, due to their toxicological effects. In the present work, the diamines putrescine and cadaverine were also investigated since they may potentiate the toxicity of the above amines, and the ight serve ax indicators of poor hy; some food substrates (Maring-Font et al., Table 3 shows the qualitative and quantitative production of BA by LAB strains. Forty percent of the LAB strains were found to be tyramine formers (100% of the carnobacteria and the Enterococcus strains studied), Accumulation of tyramine in the fermenting broth showed a wide variation, from 20 to 5000 mg/1, being higher than 400 mg/1 in most of the cases, Levels of tyramine formed were relatively high compared to those previously reported. This may be due to the composition of decarboxylase medium, the incubation time, and to the particular activity of the strains assayed. As in other studiesS. Bover-Cid. Table 3 |. Holzapfel | imerational Journal of Food Microbiology $3 (1999) 33~41 2 Reet, Quantified (mg/t broth) biogenic amine production by lactic acid bacteria strains: determination was made after cultivation at 37°C for & days in decarboxylation broth” ‘Species Surains TY" HI PU CA tested a —_— Camobacterium divergens 0 oy 10483 - = = Camobacterium gallinarum 1 ay 1504 - = : Carsobucterium piscicola 4 4) 159 - - Se Enterococcus durans 1 a) 610 - - 5 Enreracoceus faecalis 15 (15) 601-4986 - e : Emteracoceus faecium 10 (10) 379-4339 - - 5 Lactobacillus acidepitas 5 ay 498 a _ Lactobacillus buvaricus 4 2) 458-551 . - . Lactobacillus bifermenans 1 a - Ee : Lactobucillus brevis 4 o - 0) a8 a ois Lactobucillus buchneri ' a - - : Lactobacillus bulgaricus 1 ay - . : Lactobucilus case’ i a - . Lactobacillus curvatus Is aay = 37-1830 Lactabacillus dethruechii ssp. bulgaricus — 2 a ao - . Lactobucils johnson 4 a - . . Luctabucillus paracuse’ ssp. purucusei 6 a - - . Lactobacillus plantarum 16 a - 5 . Lactobacillus sakei 7 o - : S Lactococcus lactis ssp. lactis 1 a - - . Lewe, mesenteroides ssp. mesentervides 1 a) Bs - - - Leucenastoe carnosian 2 @) Not-asa0 = . . Streptococcus thermophilus 2 Cy 5304 : 5 . Lactobacillus sp. n (3) Ao-2943 - - oy * Decarboxylavion broth contained O.S% tyrosine and 0.2% of histidine, ornithine and Iysine, respectively “TY, tyramine: Hl, histamine: PU, putrescine: CA, eadaverine, “Number of strains which showed biogenic amine formation, (oosten and Northold, 1989; Leisner et al. 1994: Giraffa et al., 1995; Straub et al., 1995) Lb. brevis, 2 Lb. buchneri, Lb. curvatus, Leuconostoc carnoswmn, 3 Camobacterium spp. and Enterococcus spp. were = found to be the most intensive tyrosine decarbox- ‘Xf ylase species. Still, strains from other species also {Showed amine production. Only a few strains of ¥ lactobacilli produced putrescine and/or a small ¥ Amount of cadaverine along with tyramine. Although ‘some LAB have been reported to show intensive E histidine decarboxylase activity (Joosten and North- old, 1989; Straub et al., 1995), no significant his- % tamine production was observed in the LAB tested. In addition, several strains of lactobacilli, Leuconos- foc spp., Weissella spp.. and pediococci showed no decarboxylase activity. The results of BA production in a synthetic medium suggest that the capability 10 Produce amines might be strain dependent rather © than being related to specific species. However, this fly seems to be more common among strains of s a % g & particular species, e.g., Lb. curvatus as compared to Lb. sakei (see Table 2). Several BA-forming strains are of importance in food fermentation, Some of the positive strains might be used as starter or protective cultures without knowledge about their potential to form BA (for instance, BA-positive strains of Streptococcus ther- mophilus for yoghurt production, numerous positive Lb. curvatus strains associated with sausage fermen- tation, and strains of Ent. faecium and Ent. faecalis of pharmaceutical products). Therefore, the inability to form BA needs also to be confirmed for micro- organisms usually considered safe. Most of the Enterobacteriaceae strains assayed. produced high quantities of cadaverine and putres- cine. whereas no strain was found to produce tyramine (Table 4). Production of high amounts of histamine by enterobacteria in fish has been exten- sively reported (Taylor et al.. 1978; Chen et al., 1989; Beutling, 1996), especially as causative agent4 S. Bover-Cid, WH. Holzapfel I Intemational soumal of Food Microbiology 53 (1999) 331 Table 4 ‘Quantified mg/1 broth) biogenic amine production by Enterobacteriaceae strains: determination was made after cultivation at 37°C for 4 days in decarboxylation broth” Species Swrains TY HI PU cA 5 tested = Citrobacter freundit 1 = ay a mm 3 a 3B . Enterobacter cloacae . 2 18-29 re) 473-568 Q 599-755 Klebsiella axyioca 1 - ay 38 - a 683 Serratia liquefaciens 4 - a 1645 w 474-758 « 591-971 Serratia marcescens 1 - a 16 a 74 a 594 Serratia sp. 2 - ay 2 a a7 ay 390 Enterobacteriaceae 1 - 5 a 476 a 599 “ Decarboxylation broth contained 0.5% tyrosine and 0.2% of his See Table 3. “Number of strains which showed biogenic amine formation, of scombroid poisoning. Yet. only minor formation of histamine was observed in the present study for the enterobacteria, which were originally isolated from meat products. These results are in agreement with the fact that fermenting microorganisms (LAB) are occasionally: associated with tyramine formation, although they can also contribute to the accumulation of other BA such us putrescine, Since LAB may constitute part of the spoilage association, they may also be respon- sible for tyramine accumulation during food spoil- ge, On the other hand, Enterobacteriaceae would be associated with cadaverine, putrescine, and histamine formation in meat products, mainly when a deteriora- tion process occurs in either raw materials or end- products. The negative result for biogenic amine production in laboratory media does not imply similar behaviour in a food product. It should be considered that food products ure complex systems with a wide number of factors influencing microbial growth and_ activity. Little is known about the effect of different factors (such as glucose and NaCl concentration, additives, temperature, ete.) on BA production, This feature seems 10 be of special interest for fermenting food bacteria involved in fermentation. Furthermore. the influence of the redox potential and the shift from aerobic to anaerobic conditions on BA production in packaged commodities requires more attention. Also. interactive effects between BA and food associated microorganisms may provide another rewarding field of study. ine. ornithine and lysine. respectively. Conclusions A decarboxylase screening medium for LAB and Enterobacteriaceae has been modified. It enables improved detection of BA positive strains. The plate procedure still presents some limitations in terms of sensitivity in detecting BA forming microorganisms. However, compared to chromatographic analysis, the decarboxylase medium allows a rapid preliminary selection of strains with low decarboxylase activity, with a detection limit estimated around 350 mg/l. BA production by bacteria has thus far been most extensively studied with respect to tyramine and histamine, probably the two most important BA of acterial origin in food, In association with fer= mented foods, some LAB such as the enterococci (generally), carnobacteria, and some: strains, par- ticularly of Lb, curvans, Lb. brevis and Lb. buch neri, have importance as tyramine producers, As food contaminants, the Enterobacteriaceae probably rank first as decarboxylase active bacteria mainly producing cadaverine and putrescine, Acknowledgements The authors gratefully acknowledged the financial support received by the EU Commission (Cost-Ac- tion 917: Biologically active amines in food: and FAIR CT-97-3078) and the ‘Comisién Interminis- terial de Ciencia y Tecnologia’ (ALI95-0890-C04-Ol) of the “Ministerio de Educacién y Ciencia’, Spain. References Baranowski, J.. 1985. Assay for histamine decarboxylase activity. In: Pan, B.S.. James, D. (Eds., Histamine in Marine Products Production by Bacteria, Measurement and Pre of Forma- tion, Vol, 252. pp. 10-13, FAO Fish, Tech. Paper. 1996, In: Biogene Amine in der Ernithrung, 1990, J formation of biologically active amines in fouls. Int, J, Foon! Microbiol, 11, 73-34. Chen, CML. Wie, Cul. Koburger, J... Marshall, M.R.. 198! Comparison of four agar media for detect 1 J. Food! Protect, $4 (11), 808 ser. D.. Hammes, WP, Holzap mine and tyramine by lactic fel, W1 seid acter 11, 278-281. De Man, J.C., Rogos, M., Sharpe. M.E., 1960. A medium cultivation of lactobacilli, J, Appl. Bacteriol. 23, 120-135 Frank, HA., Baranowski Premaratne, RJ. 198 tivities of eters (Coryphaenas hippueras) after incubation at (and 32°C, Ant Food Microbiol 4. Gale, ELF, 1946, The bacterial amino acid decarbuoxylases. tn Nord, FF. (Ed, Advances in Enzymology, Vol. vol. 6, E Interscience Publishers, New York. & Ginaffa, G.. Pep G., Lucci, F., Neviani, E., Carminati, D,, 1995. 422, ,Hacmolytic activity. production of thermonuclesse and ‘Biogenic amines by dairy enterococci, Mal. 1. Food Sei. 4, i MI-349, Halisz, A., Barith, A., Simon-Sarkadi. L.. Holzapfel, W.H., 1994. Biogenic amines and their production by microorganisms in HF food. Trends Food Sci. Technol. 5, 42-9. dez-Jover, T., zquierdo-Pulido. M...Veciana-Nogués, M.T.. Vidal-Carou, M.C.. 1996. Jon pair high performance liquid carboxylase met chromatographic determination of biogenic amines im meat and meat products. J. Agric. Food Chem, $4. 2710-2715."*-" Josten, HM.LJ.. Nonhold, M.D.. 1989. Detection. growth and amine-producing capaci Environ, Microbiol. 55 (91, Karmas. E1981, Biogenic amines as indicators. of seafood freshness. Lebensm. Wiss. Technol, 14, 273=275. Leisner. JU. Millan. J.C, Huss. HH. Larsen, LM, 1994, Production of histamine and tyramine by lactic acid bacteria isolated from vacuum-packed sugar-salted fish. J. Appl. Bae- teriol, 76, 417-423. Maijala. R.. 1993. Formation of histamine and tyramine by some a bacteria in MRS-broth and mevlitied decarboxylation Appl. Microbiol, 17, 400-33 A.. Vidal-Carou, M.C., Wzquierdo-Putide, ML, Vee Noguds, M.T.. Henindez-Jover. T.. 1995, Lex amines biognes dans les aliments: leur signi ‘Aan, Pals, Exp. Chim. 88, 119140, Mueller. V.. 1984. Distribution of amino Emterobaeteriacese, Acta Pathol, Microbiol, § 277. Niven, CR. Jeffey. MR, Corletl, DA. 1981. Ditlerential plating medium for quntitativ of histamine produce ing bacteria. Appl. Environ, Mietobiol, 41, 321-322. Rees PLA.. Moore, W.M., Snell, E.E.. 185, Pyruvolyl-dependent histidine devarboxylases from Clostriliam perfringens. aul Lactobacillus buctneri 3, Biol, Chem. 258 (1), AYA, Herrero, MM. LiipezSihs fentura, M-T., 1997, Evalue and tyramine-producing bia Appl, Microbiol, 28, 309-312, Straub, BW., Kicherer, M.. Schilcher, S.M., Hammes, WP, 1995, i¢ amines by fermentation orga Z Lebensm, Unters. Forsch, 201, 79-82, Taylor, S.L., Gutherz, L.S., Leatherwood, M., Tillman, F., Lieber. E.R. 1978, Histamine porduction by (ood-bome bacterial species, J. Food Safety 1, 173-187. Vidal-Carou, M.C., I2quierdo-Pulido, M., Martin-Momro, M., Mariné-Font, A., 1990, Histamine and tyramine in meat relationship with mest spoilage. Food Chem. 37. ia in ripened sausages. Lett